Abstract: Embodiments of the present invention comprise methods and systems for automatically adjusting images to conform to preference data.

Abstract: The present invention describes a partial independent component analysis (PICA) technique for blindly separating partially independent and/or gaussian-like sources from mixed observations over an informative index subspace, which allows various applications in independent component imaging. The present invention estimates a demixing matrix using only the independent and/or nongaussian portion of the observations. Specifically, rather than using all the data points which give rise to a large separation error, a subset of the data points is identified such that the partial source profiles defined over such a subset are statistically independent and/or nongaussian. The present invention describes a complete implementation of such a technique, whose steps and parameters may be achieved and estimated using an information theoretic-based neural computational algorithm.

Abstract: A text mining program is provided that allows a user to perform text mining operations, such as: information retrieval, term and document visualization, term and document clustering, term and document classification, summarization of individual documents and groups of documents, and document cross-referencing. This is accomplished by representing the text of a document collection using subspace transformations. This subspace transformation representation is performed by: constructing a term frequency matrix of the term frequencies for each of the documents, transforming the term frequencies for statistical purposes, and projecting the documents or the terms into a lower dimensional subspace. As the document collection is updated, the subspace is dynamically updated to reflect the new document collection.

Abstract: An artificial neural network (ANN) based system that is adapted to process an input pattern to generate an output pattern related thereto having a different number of components than the input pattern. The system (26) is comprised of an ANN (27) and a memory (28), such as a DRAM memory, that are serially connected. The input pattern (23) is applied to a processor (22), where it can be processed or not (the most general case), before it is applied to the ANN and stored therein as a prototype (if learned). A category is associated with each stored prototype. The processor computes the coefficients that allow the determination of the estimated values of the output pattern, these coefficients are the components of a so-called intermediate pattern (24). Assuming the ANN has already learned a number of input patterns, when a new input pattern is presented to the ANN in the recognition phase, the category of the closest prototype is output therefrom and is used as a pointer to the memory.

Abstract: A neural sensor is provided which receives raw input data defining a pattern, such as image or sound data, and generates a classification identifier for the pattern. The neural sensor has a pattern array former which organizes the raw input data into the proper array format. A first order processing section receives the pattern array and generates a first order feature vector illustrative of first order features of the input data. A second order processing section also receives the pattern array and generates at least one second order feature vector illustrative of gradients in the input data. A vector fusion section receives the feature vectors from the first and second order processing sections and generates a single fused feature vector which is provided to a pattern classifier network. The pattern classifier network, in turn, generates a pattern classification for the input data.

Abstract: A system for determining the optimal settings for parameter of a stencil printing machine. The system generates a model mapping parameter inputs to output results. The model is then used to determine the optimal settings of parameter inputs in order to produce the desired results. One form of mapping is to generate a neural network model of a system. The neural network is generated from data that includes multiple sets of input parameter settings and the resulting output associated with the inputs. Back propagation is then performed on the neural network to determine the optimal settings.

Abstract: Embodiments of the present invention provide digital watermarking methods that embed a digital watermark in both the low and high frequencies of an image or other production, providing a digital watermark that is resistant to a variety of attacks. The digital watermarking methods of the present invention optimize the strength of the embedded digital watermark such that it is as powerful as possible without being perceptible to the human eye. The digital watermarking methods of the present invention do this relatively quickly, in real-time, and in an automated fashion using an intelligent system, such as a neural network. The digital watermarking methods of the present invention may also be used in a variety of new applications, such as the digital watermarking of sensitive aircraft parts and military equipment.

Abstract: A neural network has been optimized to function as an image preprocessor. The image processor evaluates input imagery and outputs regions of interest, ignoring backgrounds or data features that differ from programmed geometries. The smart imager algorithm has been applied to medical and military datasets. Results from over 200 patient images demonstrate that the image preprocessor can reliably isolate information of diagnostic interest in pulmonary data. Similarly, a smart preprocessor reliably locates peaks in correlation surfaces in an automated target recognition application. In both cases, the smart imager is able to ignore noisy artifacts and background information, highlight features of interest and improve detection system performance.

Abstract: Abnormal regions in living tissue are detected by obtaining images from different views of the living tissue; performing single-image CAD of each image to determine suspected abnormal regions depicted in the image; and combining measurements of the suspected abnormal regions in each image to determine whether a suspected abnormal region is an abnormal region. The living tissue may be a human breast and the abnormal region may be a mass in the breast. Ipsilateral mammographic views of the breast, a craniocaudal view, and a mediolateral oblique view may be used. Features which are relatively invariant or behave predictably with respect to breast compression are extracted using the single-image CAD and then combined.

Abstract: A control system for a harvester or similar implement includes a supervisory controller, a set of low-level controllers and a neuro-fuzzy inference system. The supervisory controller employs human expert knowledge and fuzzy logic. The controller monitors the quality of the harvesting process, such as gain loss, dockage, grain damage and the like. Based on the measurements, setpoints for all critical functional elements of the implement are determined. The neuro-fuzzy inference system determines machine settings according to operating conditions and learns from harvester experience. The parameters of the neuro-fuzzy inference system are stored in on-board memory. The neuro-fuzzy system can be used for harvester set-up and as one of the knowledge sources for repeated adjustments during the harvest.

Abstract: A fast localization with advanced search hierarchy system for fast and accurate object localization in a large search space is based on an assumption that surrounding regions of a pattern within a search range are always fixed. The FLASH system comprises a hierarchical nearest-neighbor search system and an optical-flow based energy minimization system. The hierarchical nearest-neighbor search system produces rough estimates of the transformation parameters for the optical-flow based energy minimization system which provides very accurate estimation results and associated confidence measures.

Abstract: A classification apparatus, notably for uses in recognition or characterisation of odors, comprises a plurality of sensors for generating raw data representing a plurality of instances of a plurality of different classes; and a processing unit for processing said raw data so as to determine an identification model. The identification model comprises definitions of the classes and a particular allocation rule selected dependent upon the application. The class definitions are established by analysing data obtained during a learning phase, the analysis being performed according to a particular information extraction method. During a later identification phase, the identification model enables an instance of unknown class to be allocated to an appropriate class amongst those defined during the learning phase. The information extraction method can be selected dependent upon the application.

Abstract: A spatial image processor neural network for processing image data to discriminate between first and second spatial configurations of component objects includes a photo transducer input array for converting an input image to pixel data and sending the data to a localized gain network (LGN) module, a parallel memory processor and neuron array for receiving the pixel data and processing the pixel data into component recognition vectors and chaotic oscillators for receiving the recognition vectors and sending feedback data to the LGN module as attention activations. The network further includes a temporal spatial retina for receiving both the pixel data and temporal feedback activations and generating temporal spatial vectors, which are processed by a temporal parallel processor into temporal component recognition vectors. A spatial recognition vector array receives the temporal component recognition vectors and forms an object representation of the first configuration of component objects.

Abstract: Digitized image data are input into a processor where a detection component identifies the areas (objects) of particular interest in the image and, by segmentation, separates those objects from the background. A feature extraction component formulates numerical values relevant to the classification task from the segmented objects. Results of the preceding analysis steps are input into a trained learning machine classifier which produces an output which may consist of an index discriminating between two possible diagnoses, or some other output in the desired output format. In one embodiment, digitized image data are input into a plurality of subsystems, each subsystem having one or more support vector machines. Pre-processing may include the use of known transformations which facilitate extraction of the useful data. Each subsystem analyzes the data relevant to a different feature or characteristic found within the image.

Abstract: A 2D image supplied from an image input unit including a wide view lens is sampled into a discrete form by an array sensor, and then mapped to a multi-resolution space by a 2D filter. The feature of the supplied image is detected, and then the mapped image is transformed to a local pattern about the detected feature, and then the coordinates of the position of the feature and the code word of the local pattern are formed into a set which is then encoded. The code is supplied to each cell of a stochastic automaton. The quantity of visual information is calculated in accordance with the quantity of mutual information between different cells of the stochastic automaton consisting of cells in blocks, the coordinates of the position of the feature and the distance from the feature to the optical axis so as to control the optical axis of the image input unit in such a manner that the quantity of visual information is maximized.

Abstract: Methods, code and apparatus analyze cell images to automatically identify and characterize the Golgi complex in individual cells. This is accomplished by first locating the cells in the image and defining boundaries of those cells that subsume some or all of the Golgi complex of those cells. The Golgi complex in the images typically have intensity values corresponding to the concentration of a Golgi component in the cell (e.g. a polysaccharide associated with the Golgi complex). The method/system then analyzes the Golgi components of the image (typically on a pixel-by-pixel basis) to mathematically characterize the Golgi complex of individual cells. This mathematical characterization represents phenotypic information about the cells' Golgi complex and can be used to classify cells. From this information, mechanism of action and other important biological information can be deduced.

Abstract: The present invention relates to an apparatus for rapidly analyzing frame(s) of digitized video data which may include objects of interest randomly distributed throughout the video data and wherein said objects are susceptible to detection, classification, and ultimately identification by filtering said video data for certain differentiable characteristics of said objects. The present invention may be practiced on pre-existing sequences of image data or may be integrated into an imaging device for real time, dynamic, object identification, classification, logging/counting, cataloging, retention (with links to stored bitmaps of said object), retrieval, and the like. The present invention readily lends itself to the problem of automatic and semi-automatic cataloging of vast numbers of objects such as traffic control signs and utility poles disposed in myriad settings.

Abstract: An automated method, storage medium, and system for analyzing bone. Digital image data corresponding to an image of the bone are obtained. Next there is determined, based on the digital images, a measure of bone mineral density (BMD) and at least one of a measure of bone geometry, a Minkowski dimension, and a trabecular orientation. The strength of the bone is estimated based upon the measure of BMD and at least one of the measure of bone geometry, the Minkowski dimension, and the trabecular orientation. To improve bone texture analysis, the present invention also provides a novel automated method, storage medium, and system in which digital image data corresponding to an image of the bone is obtained, and a region of interest (ROI) is selected within the bone. A fractal characteristic of the image data within the ROI using an artificial neural network is extracted. The strength of the bone is estimated based at least in part on the extracted fractal characteristic.

Abstract: A method and apparatus for color matching are provided, in which paint recipe neural networks are utilized. The color of a standard is expressed as color values. The neural network includes an input layer having nodes for receiving input data related to paint bases. Weighted connections connect to the nodes of the input layer and have coefficients for weighting the input data. An output layer having nodes are either directly or indirectly connected to the weighted connections and generates output data related to color values. The data to the input layer and the data from the output layer are interrelated through the neural network's nonlinear relationship. The paint color matching neural network can be used for, but not limited to, color formula correction, matching from scratch, effect pigment identification, selection of targets for color tools, searching existing formulas for the closest match, identification of formula mistakes, development of color tolerances and enhancing conversion routines.

Abstract: Apparatus for fraud detection for an account includes a plurality of N statistical estimators which estimate the likelihood of a transaction being fraudulent from data about a transaction on the account and transaction history of the account. A statistical estimator is provided which computes an estimate of the likelihood of a particular signature being that of a person authorized to sign on the account from an exemplar of the signature and a history of previous signatures. A combiner produces a combined probability estimate from the plurality of N statistical estimators and the statistical estimator.

Abstract: A method and an apparatus of compressing data. The method and apparatus include constructing a neural network having a specific geometry using a finite and discrete Radon transform. The data is then fed through the neural network to produce a transformed data stream. The transformed data stream is thresholded. A fixed input signal is fed back through the neural network to generate a decoding calculation of an average value. The thresholded data stream is entropy encoded.

Abstract: A document processing apparatus comprises a scanner for scanning a bank check to obtain gray scale image data associated with the bank check. A processor is provided for (i) selecting one of a plurality of binarization programs based upon the gray scale image data, and (ii) applying the binarization program selected to at least a portion of the gray scale image data to provide a binary image of at least a portion of the bank check.

Abstract: A neural network pattern recognition system for remotely sensing and identifying chemical and biological materials having a software component having an adaptive gradient descent training algorithm capable of performing backward-error-propagation and an input layer that is formatted to accept differential absorption Mueller matrix spectroscopic data, a filtering weight matrix component capable of filtering pattern recognition from Mueller data for specific predetermined materials and a processing component capable of receiving the pattern recognition from the filtering weight matrix component and determining the presence of specific predetermined materials. A method for sensing and identifying chemical and biological materials also is disclosed.

Abstract: A method and apparatus is provided which analyzes an image of an object to detect and identify defects in the object utilizing multi-dimensional wavelet neural networks.

Abstract: A method for monitoring fabrication processes of finely structured surfaces in a semiconductor fabrication includes the steps of providing reference signatures of finely structured surfaces, measuring at least one signature of a test specimen surface, comparing the measured signature with the reference signatures, and classifying the test specimen surface by using the comparison results, wherein the measurement of the reference signatures is carried out by measuring the local distribution and/or intensity distribution of diffraction images on production prototypes having a specified quality. The classification is preferably carried out here with a neural network having a learning capability and/or a fuzzy logic. Furthermore, a device for carrying out the method is provided.

Abstract: A system for automatically generating a database of images and positions of objects of interest identified from video images depicting roadside scenes that are recorded from a vehicle navigating a road and having a system that stores location metrics for the video images.

Abstract: An alphanumeric character image recognition system includes a first stage comprising at least a first, second and third digital image signal processing network having each at least one input terminal and at least one output terminal and said networks being designed to process image information from digital image signals, and comprising at least a first, second and third memory register having each at least one input terminal and at least one output terminal and the input terminals of the first, second and third memory registers being connected to the output terminal of the first network, the output of the second network and the output terminal of the third network respectively and said memory registers being designed to contain the image information processed by the first, second and third digital image signal processing networks, and a second stage characterized in that said second stage comprises at least one first and one second classifier network having each at least one first and one second input termina

Abstract: A multi-neural net imaging apparatus and method for classification of image elements, such as biological particles. The multi-net structure utilizes subgroups of particle features to partition the decision space by an attribute or physical characteristic of the particle and/or by individual and group particle classification that includes an unknown category. Preprocessing and post processing enables heuristic information to be included as part of the decision making process. Preprocessing classifies particles as artifacts based on certain physical characteristics. Post processing enables the use of contextual information either available from other sources or gleaned from the actual decision making process to further process the probability factors and enhance the decisions.

Abstract: An automated method, storage medium, and system for analyzing bone. Digital image data corresponding to an image of the bone are obtained. Next there is determined, based on the digital images, a measure of bone mineral density (BMD) and at least one of a measure of bone geometry, a Minkowski dimension, and a trabecular orientation. The strength of the bone is estimated based upon the measure of BMD and at least one of the measure of bone geometery, the Minkowski dimension, and the trabecular orientation. To improve bone texture analysis, the present invention also provides a novel automated method, storage medium, and system in which digital image data corresponding to an image of the bone is obtained, and a region of interest (ROI) is selected within the bone. A fractal characteristic of the image data within the ROI using an artificial neural network is extracted. The strength of the bone is estimated based at least in part on the extracted fractal characteristic.

Abstract: An automated guided apparatus capable of accurately determining its position within a walled environment such as a mine or building. A mobile unit incorporating an inertial measurement unit and a gray scale vision system processor/camera and/or a laser pointer is able to initialize its location and then update its location within the environment. The apparatus is especially adapted for producing tunnel plan views (“TOPES”) and also for guiding equipment through such environments.

Abstract: A neurofilter is implemented as a neural network in which the weighting coefficients have previously been set, by an appropriate training procedure, such as to provide a desired form of filter response. The neurofilter is applicable to filtering of image data or serial data signals. Also, by training a neurofilter to produce output data based on amounts of error that occur in the output data from a conventional filter, a filter apparatus can be provided whereby the neurofilter compensates for errors in output data from the conventional filter. The design and manufacturing constraints on the conventional filter can thereby be substantially relaxed.

Abstract: In an image processing method and apparatus, image data having multi-value levels for one pixel is input, and the input image data is quantized such that an output area of one pixel is adapted to an output device in which an output area of one pixel changes depending on the position of the pixel. A quantizing process executes an arithmetic operation based on an algorithm of a neural network on the basis of a value obtained by multiplying an output value by a weight corresponding to an area of each pixel. Therefore, even if pixels have different maximum luminances, the different numbers of bits, and different color expression capabilities, an optimum half-tone process can be performed by an algorithm based on a cellular neural network, and a high-quality image can be obtained.

Abstract: The present invention relates to a hierarchical artificial neural network (HANN) for automating the recognition and identification of patterns in data matrices. It has particular, although not exclusive, application to the identification of severe storm events (SSEs) from spatial precipitation patterns, derived from conventional volumetric radar imagery. To identify characteristic features a data matrix, the data matrix is processed with a self organizing network to produce a self organizing feature space mapping. The self organizing feature space mapping is processed to produce a density characterization of the feature space mapping. The self organizing network is preferably completely unsupervised. It may, under some circumstances include a supervised layer, but it must include at least an unsupervised component for the purposes of the invention. The “self organizing feature space” is intended to include any map with the self organizing characteristics of the Kohonen Self Organizing Feature Map.

Abstract: A method and apparatus for managing at least part of a communications network and particularly for managing a customer network, for example, within an asynchronous transfer mode communications network. A predictor is used to predict parameters such as bandwidth levels and to predict when the parameter(s) will exceed capacity or previously agreed thresholds. These agreed levels may be specified for example in a service level agreement between a service provider and a customer. The predictor also predicts, how much excess there will be and how long this will occur for. This information is provided to the service provider/customer and also can be provided to an agent which comprises a computer system. This agent negotiates on behalf of either the customer (for example) with another agent acting on behalf of the service provider (for example) and in this way new terms for an agreement between the two parties is obtained.

Abstract: A method for estimating scenes from images generates a plurality of scenes and renders an image for each scene. The scenes and corresponding images are partitioned into patches. Each patch is quantitized as a vector and probability density function is fitted to each vector. The patches and probability densities are organized as a Markov network where local probability information is propagated to neighboring nodes. After propagation, the probability density at each node is used to estimate the scene.

Abstract: An information processing apparatus including a switch for manually requesting change of output image quality, detecting unit for detecting a condition of the apparatus, setting unit of setting an image forming condition in accordance with the detected condition and the requested output image quality and control unit for changing the image forming condition by learning the request previously requested by the user. Image forming conditions are adjusted so as to satisfy the user manually based upon the degree of satisfaction determined by the user.

Abstract: An image processing apparatus of the present invention is provided with a region separating section and an automatic adjusting section. The region separating section, upon receiving a reference document having a specified ratio of a character region, a photographic region, and a spot region, recognizes and separates the each region of the received document from one another. The automatic adjusting section (1) counts the number of pixels in each of the separated regions and (2) uses the number of pixels thus counted to change a density conversion table, a filter, or a region separation table so as to adjust the separated state of each region such that the number of pixels thus counted is substantially equal to the number of pixels of the region separation state, thereby improving the accuracy of region separation of the image processing apparatus and the image quality of a final image after it is processed with image processing.

Abstract: The method provides object recognition procedure and a neural network by using the discrete-cosine transform (DCT) (4) and histogram adaptive quantization (5). The method employs the DCT transform with the added advantage of having a computationally-efficient and data-independent matrix as an alternative to the Karhunen-Loeve transform or principal component analysis which requires data-independent eigenvectors as a priori information. Since the set of learning samples (1) may be small, we employ a mixture model of prior distributions for accurate estimation of local distribution of feature patterns obtained from several two dimensional images. The model selection method based on the mixture classes is presented to optimize the mixture number and local metric parameters. This method also provides image synthesis to generate a set of image databases to be used for training a neural network.

Abstract: An image-processing method is provided with the first step of dividing an input image having n(n>1) gray scales into a plurality of matrixes, the second step of carrying out at least either a resolution-converting process or a variable magnification process for each of the divided matrixes, by using a hierarchical neural network that can execute a learning process for each input image, and the third step of outputting the image processed in the second step as an output image having n gray scales. Thus, the weights adjustment of the network can be carried out on each input image whatever image is inputted thereto; therefore, it is possible to always provide an optimal converting process.

Abstract: A system, method, and software product provide for unsupervised identification of complex, nonlinear subspaces in high dimensional data. The system includes a vector quantization module, a weighted topology representing graph module, and an encoding module. The vector quantization module takes vector data inputs and extracts a group of inputs about a number of cluster centers, using a globally optimized clustering process. The weighted topology representing graph module creates a weighted graph of the vector space, using the cluster centers as nodes, weighting edges between nodes as a function of the density of the vectors between the linked nodes. The encoding module uses the weighted graph to recode the input vectors based on their proximity to the cluster centers and the connectedness of the graph. The recoded vectors are reinput into the vector quantization module, and the process repeated until termination, for example at a limited number of cluster centers.

Abstract: A method and apparatus enhance visible contrast within an acquired image for display. The contrast enhancement utilizes all N bands of an original N-band spectral image to produce an M-dimensional enhanced image for display. The method creates an enhanced image from an original image in which the visible contrast in the original image is improved. The original image includes pixels, each pixel having N spectral intensities. The display or printer device which must display the relevant information may be limited to a number of bands M which is smaller than N. Maximum contrast of objects is obtained by emphasizing differences in the N-dimensional pixels by as large differences as possible within the dynamic range of the M-band display space. When M=N=3, this means moving pixels in the display space to utilize the full color palette available on a color monitor or printer. When N>M, a mapping from N space to M space must also be accomplished.

Abstract: A method and an arrangement are presented for pattern recognition on the basis of statistics. According to the method, for an object to be recognized on the basis of a complete set of two-class or multiclass classifiers, the association with each target class of the class set is estimated with a numerical value that is produced by cascaded use of polynomial classifiers. According to the invention, on a learning sample in which all class patterns to be recognized are sufficiently represented, there is a selection, from all the two-class or multiclass classifiers by way of their estimation vector spectrum, of those two-class or multiclass classifiers with estimations contributing the most to minimize a scalar quantity calculated over the estimation vector spectrum and having high separating relevance. The selected two-class or multiclass classifiers are subsequently used to form, via an expanded learning sample, estimation vectors from which expanded characteristic vectors are produced by polynomial linking.

Abstract: The computer imaging analysis of bakery products for quality control and other purposes is disclosed. Apparatus and methods useful in such analysis are disclosed. The methods are useful on all types of bakery products. They can be used to analyze for parameters, such as size, shape, area and volume. They can also be used to analyze holes, grain or crust. Viewable images may be provided. The parameters can be compared with prescribed specifications and can be used on a plurality of products to determine substantial uniformity.

Abstract: There is disclosed a pattern identifying neural network comprising at least an input and an output layer, the output layer having a plurality of principal nodes, each principal node trained to recognize a different class of patterns, and at least one fuzzy node trained to recognize all classes of patterns recognized by the principal nodes but with outputs set out at levels lower than the corresponding outputs of the principal nodes.

Abstract: An adaptive hierarchical neural network based system with online adaptation capabilities has been developed to automatically adjust the display window width and center for MR images. Our windowing system possesses the online training capabilities that make the adaptation of the optimal display parameters to personal preference as well as different viewing conditions possible. The online adaptation capabilities are primarily due to the use of the hierarchical neural networks and the development of a new width/center mapping system. The large training image set is hierarchically organized for efficient user interaction and effective re-mapping of the width/center settings in the training data set. The width/center values are modified in the training data through a width/center mapping function, which is estimated from the new width/center values of some representative images adjusted by the user.

Abstract: In a method for separating a desired color region from a color image, color component values, which represent color components of the image, are calculated. Energy minimization is then carried out, in which calculations of an update rule for minimizing energy are iterated, in accordance with an energy function, which is defined by the color component values and a line process representing the presence or absence of continuity of the color component values in the image. A contour of the color region, which contour is represented by the line process and obtained from the energy minimization, is then extracted. The desired color region is thereby separated accurately from the color image without being affected by a background in the image.

Abstract: Taking into consideration the disadvantage that a large-scale analog neural network cannot be constructed as an LSI and, even if this were possible, the cost would be prohibitive and the network would lack universality, a digital image processor for processing input image data based upon a cellular neural network is provided with a first multiply-and-accumulate arithmetic unit for digitally processing multiplication and accumulation of input image data of a plurality of pixels and input weighting values in a predetermined area, a second multiply-and-accumulate arithmetic unit for digitally processing multiplication and accumulation of output image data of a plurality of pixels and output weighting values in a predetermined area, and a non-linear acting unit for deciding output image data in accordance with results of calculation from the first and second multiply-and-accumulate arithmetic unit and non-linear characteristic parameters.

Abstract: An image reader having a shading correction function includes a scanner for scanning an original, a converter for converting the received light from the original into electric signals and outputting the electric signal as data of an read image, and a corrector for carrying out a shading correction operation on the data of the read image on the basis of a shading correction value and outputting corrected image data. The corrector computes shading correction values for the whole reading face of the original platen from the shading correction value at an arbitrary position on the reading face and carries out the shading correction operation on the data of the read image on the basis of the computed shading correction values.

Abstract: A method and apparatus is described for improving the efficiency of any machine that uses an algorithm that maps to a higher dimensional space in which a given set of vectors is used in a test phase. In particular, reduced set vectors are used. These reduced set vectors are different from the vectors in the set and are determined pursuant to an optimization approach other than the eigenvalue computation used for homogeneous quadratic kernels. An illustrative embodiment is described in the context of a support vector machine (SVM).

Abstract: A method for detecting a face in an image includes the steps of applying the image to a first classification tool that determines a rotational angle for rotating the image that makes the image most resemble an upright face; rotating the image by the rotational angle determined by the first classification tool; and applying the rotated image to a second classification tool, which determines whether the rotated image represents a frontal face or not.